Heterogeneous photocatalysis on metal oxide semiconductor particles has been shown to be an effective means of removing organic and inorganic pollutants from water and air streams. The effects of photocatalyst composition, structure, crystal size, band gap energy, incident light intensity, and the nature of the electron donors and acceptors on photocatalytic reactivity have been investigated.
Photochemically-excited semiconductor particles can catalyze the reduction and/or oxidation of a variety of chemical species. Charge-pair generation is achieved by the absorption of a photon with an energy greater than or equal to its band gap energy. The absorbed photon promotes an electron from the valence band into the conduction band, and in doing so, creates a positively charged valence band hole. Excited-state electrons and holes can either: (i) recombine with the release of heat; (ii) migrate through the lattice structure to various trapped sites; or (iii) migrate to the particle surface and participate in electron transfer reactions as reported by Hoffmann, et al., Chem. Rev., 1995, 95, 69. Surface-trapped conduction band electrons can reduce surface-bound oxidants (+0.5 V to -1.5 V vs NHE depending on the semiconductor and pH), while valence-band holes are capable of oxidizing a wide range of electron donors (+1.0 V to +3.5 V vs NHE).
Hydroxyl radicals, formed from the oxidation of vicinal water or bound hydroxyl groups, are generally believed to be the principal reactive species responsible for the photo-oxidation of organic compounds in semiconductor photocatalysis. In some cases, direct hole transfer (h.sup.+.sub.vb) via a photo-Kolbe mechanism has also been shown to be an important oxidation pathway. Surface-bound OH.cndot. radicals are produced by the oxidation of titanol groups by valence band holes as follows: EQU &gt;Ti(IV)OH+h.sub.vb.sup.+ .fwdarw.Ti(IV)OH.sup..cndot.+ (1)
Direct electron transfer (i.e., hole transfer) may occur as follows: ##STR1##
Charge neutrality of the particle is maintained by concurrent reduction via conduction band electrons. In an aerated system, oxygen is reduced and forms superoxide and/or hydroperoxyl radicals. These species can be further reduced to form hydroxyl radicals. ##STR2## In addition to oxygen, other species present in the liquid phase (e.g., CCl.sub.4) may be reduced if the reduction potential of the species is more positive than the conduction band energy.
Practical application of metal oxide semiconductors as photocatalysts often requires immobilization of the photocatalyst in a fixed-bed reactor configuration that allows the continuous use of the photocatalyst for treating aqueous or gaseous effluent streams by eliminating the need for post-process filtration. In conventional fixed-bed reactors, the photocatalyst is coated on the walls of the reactor, on a support matrix, or around a casing containing the light source. However, these configurations present several drawbacks such as low light utilization efficiencies due to absorption and scattering of the light by the reaction medium and restricted processing capacities due to mass transport limitations. A novel approach to solving these problems employing optical fibers as a means of light transmission and distribution to solid-supported photocatalysts was first proposed by Ollis and Marinangeli in their papers, AIChE J., 1977, 23, 415 and AIChE J., 1980, 26, 1000. However, they predicted that photocatalytic optical fiber reactor systems would not be practical due to catalyst deactivation caused by heat buildup in a bundled array. Hofstadler, et al. in their paper reported at Environ. Sci. Tech., 1994, 28, 670, have shown recently that TiO.sub.2 -coated quartz fiber rods fixed in a tubular reactor configuration using a light source adjacent to the reactor can be used to carry out oxidation of 4-chlorophenol. They reported effective quantum yields of 0.0002. In a 1991 masters thesis by D. C. Gapen, "Photocatalytic Degradation of Chlorinated Hydrocarbons", Univ. of Wisconsin, use of an optical fiber cable reactor (hereinafter "FOCR") for photocatalytic oxidation was investigated. Problems with delamination of the TiO.sub.2 coating resulted in partial degradation of 3-chlorosalicylic acid. However, a maximum quantum efficiency of 0.069 for the partial photolytic and photocatalytic oxidation of 3-chlorosalicylic acid was estimated.
Optical fibers in an FOCR can be treated as elementary waveguides having a core with an outer reactive layer. In this case, light in the form of a plane wave propagating in the core medium (designated by the subscript "1") having a refractive index n.sub.1, and which is incident on an interface with the outer medium (designated by the subscript "2") having a refractive index n.sub.2, is partially reflected and partially refracted. Snell's Law states that the ratio of the sine of the wave's incident angle, .theta..sub.i, and the sine of the wave's refracted angle, .theta..sub.refr, is constant. This ratio is equal to the ratio of the wave's velocities in the respective mediums and the inverse ratio of the refractive indices according to the formula, ##EQU1##
For fibers in which n.sub.1 is higher than n.sub.2, the propagating wave will be reflected within the fiber with 100% efficiency when the angle of incidence exceeds the critical angle. This critical angle, .theta..sub.c, is defined as the arcsine of the ratio of the refractive indices of the core and outer mediums. Some refraction will occur if the angle of incidence is less than the critical angle. However, if n.sub.1 is less than n.sub.2, then refraction will occur for all incident angles.
If the interface of the core and outer mediums absorbs the incident light with an absorption coefficient .alpha., in units of inverse length, n.sub.2 is replaced with a complex refractive index, n.sub.2 : EQU n.sub.2 =n.sub.2 (1-i.kappa..sub.2) (7)
The attenuation index, .kappa., is related to .alpha. by EQU n.kappa.=.alpha.c/4 .pi.v (8)
where c and v are the velocities of the light wave in a vacuum and in the medium, respectively.
For all wavelengths, the index of refraction of a single TiO.sub.2 rutile crystal is greater than that for quartz, which is constant at about 1.4 (personal communication between the inventors and R. H. French of DuPont Co., 1994). Values for n for wavelengths that are not absorbed by TiO.sub.2 (.lambda.&gt;375 nm) are constant at about 2.8. As the wavelength approaches the absorption onset of the photocatalyst, the index of refraction of the TiO.sub.2 crystal increases, reaching 5.5 at about 300 nm. Values for .kappa. similarly increase from 0 to about 1.5. As a consequence, incident light will be refracted to some degree for all incident angles. The ratio of the energy transmission to the TiO.sub.2 coating to the incident radiation at each reflection depends on the incident angle, the wavelength, and the refractive indices according to the Fresnel equations: ##EQU2## where E.sub.r and E.sub.i are the energies of the refracted and incident radiation, respectively, and the subscripts .perp. and // refer to the transverse electric and magnetic (TE and TM, respectively) polarizations of the incident light. The refraction for perpendicular polarizations monotonically decreases from a given value at normal incidence (0.degree.) to zero at the grazing angle, parallel to the quartz-coating interface. The refraction for parallel polarizations increases as the incident angle increases, reaching a maximum of 100% at Brewster's angle, (.theta..sub.B =tan.sup.-1 n.sub.1 /n.sub.2), and then falls sharply to zero as the incident angle approaches the grazing angle. The shapes of these curves are a function of the indices of refraction for quartz and the TiO.sub.2 particles (complex) and the extinction coefficient of the photocatalyst. Light incident upon a quartz-water or quartz-air interface, representative of a coating pore interface, will be totally internally reflected for incident angles greater than the corresponding critical angle with no refracted transmission of energy because the refractive index of the quartz core is greater than that of air and water. This assumption is an oversimplification of the system since the wavelength is probably at least an order of magnitude greater than the interfacial pore surface area. In this case, the interaction between the pore interface and the photon is a function of the "effective" refractive index of the pore interface. In the 1986 work by W. M. Bruno ("Powder Core Dielectric Wave Guides," Caltech Thesis, Pasadena, 1986) it has been postulated that for a powder material composed of particles, which are small compared to the incident wavelength, the refractive index of the material will be a statistical composite of the refractive indices of the particles and the void material (e.g., H.sub.2 O). In addition, the absorption of the material increases significantly as the packing fraction of the powder increases.